Linux seti_boinc compilation blues...


log in

Advanced search

Message boards : Number crunching : Linux seti_boinc compilation blues...

Author Message
Chris Bosshard
Send message
Joined: 5 Jun 99
Posts: 86
Credit: 3,474,583
RAC: 0
Switzerland
Message 9146 - Posted: 17 Jul 2004, 22:40:31 UTC

Hello again

After some helpful hints on this board I managed to compile the boinc client sucessfully. Unfortunately I'm experiencing problems compiling the seti_boinc client...

I use two Linux systems with the kernel 2.5.x
The boinc and seti_boinc sources are the latest daily builds. The configure and make of the boinc client works flawlessly.

The configure of seti_boinc works as well without any error messages. When running the make command I get the following messages:

c/image_libs/ -I.. -I../db -c -o analyzeReport.o analyzeReport.cpp
analyzeReport.cpp: In function `int ReportPulseEvent(float, float, float, int,
int, float, float, float*, int)':
analyzeReport.cpp:337: error: `assert' undeclared (first use this function)
analyzeReport.cpp:337: error: (Each undeclared identifier is reported only once
for each function it appears in.)
make[1]: *** [analyzeReport.o] Fehler 1
make[1]: Leaving directory `/home/chris/boinc/source_for_compilation/seti_boinc/client'
make: *** [all] Error 2

This happens on both my Linux boxes, one a dual processor box, the other an older P3.

Any input is appreciated.


Chris Bosshard
astroinfo Team
http://www.astroinfo.org

morthound
Send message
Joined: 9 Oct 99
Posts: 7
Credit: 361,901
RAC: 0
Sweden
Message 9149 - Posted: 17 Jul 2004, 23:03:13 UTC - in response to Message 9146.
Last modified: 17 Jul 2004, 23:27:33 UTC

Open /home/chris/boinc/source_for_compilation/seti_boinc/client/analyzeReport.cpp in an editor and add:
#include (assert.h)
Change the ( and ) for the less than and greater than sign (for some reason they don't show up on the board...)

You probably need to add it to lcgamm.cpp too.

morthound
Send message
Joined: 9 Oct 99
Posts: 7
Credit: 361,901
RAC: 0
Sweden
Message 9150 - Posted: 17 Jul 2004, 23:03:43 UTC - in response to Message 9146.
Last modified: 17 Jul 2004, 23:04:10 UTC

Chris Bosshard
Send message
Joined: 5 Jun 99
Posts: 86
Credit: 3,474,583
RAC: 0
Switzerland
Message 9307 - Posted: 18 Jul 2004, 8:13:05 UTC - in response to Message 9149.

> Open
> /home/chris/boinc/source_for_compilation/seti_boinc/client/analyzeReport.cpp
> in an editor and add:
> #include (assert.h)
> Change the ( and ) for the less than and greater than sign (for some reason
> they don't show up on the board...)
>
> You probably need to add it to lcgamm.cpp too.
>

Hi Morthound

Your input brought me a step further it starts to compile and then stops with the following error:

e_for_compilation/boinc/api -I/home/chris/boinc/source_for_compilation/boinc/lib -I/home/chris/boinc/source_for_compilation/boinc/image_libs/ -I.. -I../db -c -o version.o version.cpp
version.cpp:22: error: Syntaxfehler before `;' token
make[1]: *** [version.o] Fehler 1
make[1]: Leaving directory `/home/chris/boinc/source_for_compilation/seti_boinc/client'
make: *** [all] Fehler 2

Thanks again!


Chris Bosshard
astroinfo Team
http://www.astroinfo.org

Jbf3
Send message
Joined: 16 Jul 99
Posts: 6
Credit: 1,349
RAC: 0
United States
Message 9548 - Posted: 19 Jul 2004, 1:32:34 UTC - in response to Message 9146.

I had the same problem compiling for my Mac OSX machine. I found this thread to be usefull.

http://setiweb.ssl.berkeley.edu/forum_thread.php?id=811

The spacing in the thread is not exact, but there is a line higher up in diagnostics.h that has the proper spacing. Just copy/paste that and you should get rid of the `assert' errors.


> Hello again
>
> After some helpful hints on this board I managed to compile the boinc client
> sucessfully. Unfortunately I'm experiencing problems compiling the seti_boinc
> client...
>
> I use two Linux systems with the kernel 2.5.x
> The boinc and seti_boinc sources are the latest daily builds. The configure
> and make of the boinc client works flawlessly.
>
> The configure of seti_boinc works as well without any error messages. When
> running the make command I get the following messages:
>
> c/image_libs/ -I.. -I../db -c -o analyzeReport.o analyzeReport.cpp
> analyzeReport.cpp: In function `int ReportPulseEvent(float, float, float,
> int,
> int, float, float, float*, int)':
> analyzeReport.cpp:337: error: `assert' undeclared (first use this function)
> analyzeReport.cpp:337: error: (Each undeclared identifier is reported only
> once
> for each function it appears in.)
> make[1]: *** [analyzeReport.o] Fehler 1
> make[1]: Leaving directory
> `/home/chris/boinc/source_for_compilation/seti_boinc/client'
> make: *** [all] Error 2
>
> This happens on both my Linux boxes, one a dual processor box, the other an
> older P3.
>
> Any input is appreciated.
>
>
> Chris Bosshard
> astroinfo Team
> http://www.astroinfo.org
>

Jbf3
Send message
Joined: 16 Jul 99
Posts: 6
Credit: 1,349
RAC: 0
United States
Message 9551 - Posted: 19 Jul 2004, 1:43:21 UTC - in response to Message 9307.

The version.cpp error seem to be caused by a fault in config.h in the seti_boinc directory.

Near the bottom of the file the line:
#define VERSION_MINOR
should have an integer value after it.

Change it to something like:
#define VERSION_MINOR 0
and try again.

> Hi Morthound
>
> Your input brought me a step further it starts to compile and then stops with
> the following error:
>
> e_for_compilation/boinc/api
> -I/home/chris/boinc/source_for_compilation/boinc/lib
> -I/home/chris/boinc/source_for_compilation/boinc/image_libs/ -I.. -I../db -c
> -o version.o version.cpp
> version.cpp:22: error: Syntaxfehler before `;' token
> make[1]: *** [version.o] Fehler 1
> make[1]: Leaving directory
> `/home/chris/boinc/source_for_compilation/seti_boinc/client'
> make: *** [all] Fehler 2
>
> Thanks again!
>
>
> Chris Bosshard
> astroinfo Team
> http://www.astroinfo.org
>

Chris Bosshard
Send message
Joined: 5 Jun 99
Posts: 86
Credit: 3,474,583
RAC: 0
Switzerland
Message 10003 - Posted: 19 Jul 2004, 20:31:04 UTC

Hi Jbf3

This got me through the compilation process.
Well I have not much of an idea how to troubleshoot a failing compilation but the crunchers are helpful bunch... :-)
I will test the client and see if it works...

Unfortunately I cannot get work units for this new client.
So there is nothing left but to wait.

Thanks again!


Chris Bosshard
astroinfo Team
http://www.astroinfo.org

Darren
Volunteer tester
Avatar
Send message
Joined: 2 Jul 99
Posts: 259
Credit: 277,051
RAC: 6
United States
Message 10008 - Posted: 19 Jul 2004, 20:47:48 UTC - in response to Message 10003.


> Unfortunately I cannot get work units for this new client.
> So there is nothing left but to wait.

If you want to go ahead and test the seti executable during the wu shortage, you can do that with the test work unit included.

You can copy the file reference_work_unit.sah from the ~/seti_boinc/client/test_workunits directory to the ~seti_boinc/client directory, then rename it to work_unit.sah.

After it's in the same directory and renamed, just execute the setiathome...gnu executable with the flag -standalone. This will run it with the test work unit. Not good for any points or anything, but at least it will let you test your compile and make sure it works and then you'll at least know you're set to go when real work units are available again.


Chris Bosshard
Send message
Joined: 5 Jun 99
Posts: 86
Credit: 3,474,583
RAC: 0
Switzerland
Message 10025 - Posted: 19 Jul 2004, 21:38:01 UTC

Thanks Darren

I am already running the test workunit.
I just stumbled over the instruction after writing my last posting...

So far no errors.... now im just waiting for the result. ;-)



Chris Bosshard
astroinfo Team
http://www.astroinfo.org

Message boards : Number crunching : Linux seti_boinc compilation blues...

Copyright © 2014 University of California